AA-ending codon-based computational analyses uncover novel cytoplasmic effectors in the Magnaporthe oryzae secretome

Phytopathogens secrete diverse, sequence-unrelated effector proteins outside (apoplastic) or into (cytoplasmic) plant cells to suppress host defenses and cause devastating diseases. Effectors rapidly evolve to avoid counter-detection by the host, and fungal cytoplasmic effectors lack recognizable host cell-targeting motifs, impeding cataloguing of the full suite of host-deployed effectors encoded in pathogen genomes. In the devastating rice blast fungus Magnaporthe oryzae, cytoplasmic and apoplastic effectors are secreted by different routes, but candidates must be confirmed by fluorescent labelling, consequently few are known. AA-ending codon usage rates are the fraction of AA-ending codons as a percentage of the total number of AA- and synonymous AG-ending codons in an mRNA. In M. oryzae, high AA % rates of the corresponding mRNAs distinguished seven experimentally confirmed cytoplasmic effectors from four experimentally confirmed apoplastic effectors, but whether AA % rates can identify new effectors is unknown. Here, using computational analyses and live-cell imaging confirmation, we successfully predicted, based on AA % rates, two new cytoplasmic effectors and two apoplastic effectors in the M. oryzae secretome. Our findings support the notion that cytoplasmic effector mRNAs are globally enriched for AA-ending codons, aiding effector discovery and the search for novel sources of durable crop resistance.